Google has rolled back a recently launched Google Earth feature that let users generate AI-altered satellite, aerial and 3D images from text prompts. The tool could create plausible flyover imagery of sensitive events and locations, raising concerns that it could be used to spread misinformation despite Google’s watermarking and verification measures.

Researcher Henk van Ess demonstrated images depicting refugees near the Mexican border and a bomb crater by a Gaza hospital. Google said images carried its SynthID watermark and could be checked through Gemini or Google Lens, while van Ess recommended comparing them with independent satellite services such as Sentinel-2 and Landsat; he also showed an AI-altered Google Earth video could fool Hive’s detector.